Skip to main content
Archived

This project has been archived by its maintainers, and is no longer receiving any updates.

anyeval

Universal Evaluation for Gen AI

Usage

pip install anyeval
# run evaluation for a parquet file or directory.
anyeval run [parquet_file|dir]

Design

Data Schema (Parquet)
id dataset label input output created_at updated_at
Columns

id: is a UUID column. dataset: is a String column. label: is a JSON column. key value must be String.
input: is a JSON column. key must be String. can include media data.
output: is a JSON column. key must be String. can include media data.

media data key must start by '@[file_type]->key'
only @image->, @video->, @audio-> file types are supported.

Example:

{
  "prompt": "make the image to video",
  "@image->input_image": "fs://path/to/image.jpg"
}
{
  "inference_time": "1.234",
  "inference_device": "NVIDIA GeForce RTX 4090",
  "@video->output_video": "fs://path/to/output.mp4",
  "@video->output_video_thumbnail": "s3://path/to/output.jpg"
}

created_at: is a timestamp column.
updated_at: is a timestamp column.

Metadata

Release files for anyeval 0.1.6

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for anyeval 0.1.6
File Size Uploaded
anyeval-0.1.6.tar.gz 944.3 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for anyeval 0.1.6
File Interpreter ABI Platform
anyeval-0.1.6-py3-none-any.whl Python 3 none any Details

Total release size: 1.9 MB

Release files / anyeval-0.1.6.tar.gz

Download URL anyeval-0.1.6.tar.gz
Size 944.3 kB
Tags Source
SHA-256 checksum
How to use checksums
f5142e6c9e45601c139d8b7393f3fee52f36e61eb5490b1b7bf013045d167d8f
BLAKE2b-256 checksum
How to use checksums
43ef2640f28241e2af914b531e32c769c78fd0e2346926160225dde11a0df018
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via uv/0.7.13

Release files / anyeval-0.1.6-py3-none-any.whl

Download URL anyeval-0.1.6-py3-none-any.whl
Size 952.4 kB
Tags Python 3
SHA-256 checksum
How to use checksums
d6e36076d18879c50c750ecd9aa39162186e88881d39d48c3175fc59bcdd7f21
BLAKE2b-256 checksum
How to use checksums
2149a23a919d80dba0ad5aabbabfff2faabd69344930f8da9dedcce97ca656a6
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via uv/0.7.13
Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page